Skip to content

Pre-Built AI Agent Samples

Reference AI agent samples that use this repo's MCP Server over HTTP.

List of References
Table of Content
Sample Industry Agents Complexity
Healthcare Multi-Agent Healthcare 5 Advanced
Simple Query Agent Any 1 Beginner
Retail Shopping Assistant Retail 6 Intermediate
Financial Advisor Finance 4 Intermediate
Manufacturing Monitor Manufacturing 3 Intermediate
Education Student Assistant Education 3 Intermediate
Logistics Tracker Logistics 3 Intermediate
Insurance Claims Agent Insurance 4 Intermediate
Hospitality Concierge Hospitality 3 Intermediate
Energy Usage Advisor Energy 3 Intermediate
Real Estate Portfolio Manager Real Estate 3 Intermediate
  • healthcare-multi-agent: an advanced orchestrated multi-agent sample.
  • The other industry samples: lightweight CLIs that (a) route to a role and (b) run a search tool via MCP, optionally using openai_chat_completion for routing and summaries when available.

Key Features

  • MCP HTTP integration: Calls /health, /mcp/tools, and /mcp/execute
  • Optional LLM routing: Uses openai_chat_completion when the server exposes it
  • Search-first flow: Uses search_semantic (preferred) or search_documents
  • Minimal dependencies: requests + python-dotenv

Patterns:

  • Lightweight HTTP samples (most folders):
User
    ↓
(optional) openai_chat_completion  → role routing
    ↓
search_semantic / search_documents → retrieve relevant items
    ↓
(optional) openai_chat_completion  → concise summary
  • Advanced orchestration (healthcare-multi-agent): For example, the healthcare sample demonstrates a richer orchestrator and multi-agent handoffs.

Customization Guide

  1. Add New Industry: Copy sample, modify agent definitions
  2. Adjust Model Routing: Edit model_router.py complexity rules
  3. Add Agents: Extend agent registry with new specialists
  4. Change MCP Tools: Update agent tool permissions

Next Steps